On Fri, 2004-02-13 at 14:45, Major Variola (ret) wrote:
[sent to al-q; does cypherpunks-moderated@minder.net forward there?]
At 06:25 PM 2/12/04 -0500, Riad S. Wahby wrote:
Among others, /. is reporting that Win2k and WinNT source code may have leaked. ... If you didn't steal it, its not your problem if you read it. ... How would M$ show that you had in fact read the code? And if you didn't copy it, why would a court care?
Eric is correct in his reply to MV's article. Joe Programmer isn't necessarily obligated not to look at leaked trade secrets, but if he implements anything remotely related to the leaked secret, he and his employers or customers are subject to being sued for using the secret. In principle they can prove that the secret didn't have any influence on the work, but in practice they're stuck having to prove a negative. Eric is also correct about the "Chinese wall" between the people who look at or figure out the secret and the implementation group. Back in the early days of the IBM PC, IBM released the source to the BIOS, figuring they could use copyright to keep anyone else from implementing a compatible BIOS. Compaq did the Chinese wall trick, with one group developing a rigorous spec from the released code, then throwing it over the wall to the implementation group. If push came to shove, the implementors could have sworn that they had never seen the IBM code.